Transfer Learning for Human & AI

We propose a technique to reduce false positive detections made by aneural network using an SVM classifier trained with features derived from theuncertainty map of the neural network prediction . We demonstrate the effectiveness of this method for the detection of liver lesions on a dataset ofabdominal MR images . We find that the use of a dropout rate of 0.5 produces the least number of false positives in neural network predictions and thetrained classifier filters out approximately 90% of these false positivesdetections in the test-set .

Author(s) : Ishaan Bhat, Hugo J. Kuif, Veronika Cheplygina, Josien P. W. Pluim

Links : PDF - Abstract

Code :


Keywords : false - network - detection - positives - liver -

• Bioinformatics Specialization on Coursera
• School of Data Science
• Open Machine Learning Course
• Data Scientist with Python
• Mathematics for Machine Learning by Marc Peter Deisenroth, A. Aldo Faisal and Cheng Soon Ong
• Introduction to TensorFlow for AI, DL and ML by Andrew Ng
• Kaggle Learning
• AI Programming with Python